The text of the Torahתּוֹרָה
Hebrew text: Westminster Leningrad Codex
Status: Public domain
Consonants, vowels (niqqud), cantillation accents (te’amim), maqaf, paseq and the signs פ and ס of the open and closed parashot, following the Leningrad Codex (c. 1008) in the Westminster Leningrad Codex digital edition, in the version of the Open Scriptures Hebrew Bible (see below). Where the text has a ketiv and a qere, Toraoke shows the qere, that is, the form that is read.
The text of the Westminster Leningrad Codex is in the public domain.
Open Scriptures Hebrew Bible
Lemmas and grammatical analysis: Open Scriptures Hebrew Bible
Status: Open license
The lemma of each word (with its Strong’s number) and the morphological analysis from which Toraoke derives the part of speech, gender, number, person, state, conjugation and suffixes shown in the “Grammar” panel. The correspondence between the Hebrew verse numbering and that of the translations also comes from here.
Original work of the Open Scriptures Hebrew Bible available at https://github.com/openscriptures/morphhb. License CC BY 4.0; the Italian labels are Toraoke’s.
openscriptures/morphhb · hb.openscriptures.org
Strong’s dictionary: Open Scriptures
Status: Open license
A Concise Dictionary of the Words in the Hebrew Bible by James Strong (1894), in the Open Scriptures digital edition (electronic text by David Instone-Brewer, corrections and transliterations by David Troidl). Toraoke uses its dictionary form (lemma) of each word, in the “Grammar” panel.
Copyright 2010 Open Scriptures, Creative Commons BY-SA license (the version is not stated by the source). The original 1894 dictionary is in the public domain.
openscriptures/strongs
Italian translation: La Sacra Bibbia, versione Riveduta (1927)
Status: Public domain
The revision of Giovanni Diodati’s Bible carried out under the direction of Giovanni Luzzi (died 1948) and published in 1927, in the digital text distributed by eBible.org with the contribution of the Società Biblica in Italia. It is a Christian (Protestant) translation and numbers the verses as Christian Bibles do: Toraoke matches it to the Hebrew verses. In 27 verses the digital text attributed to the preceding verse the words that introduce a speech (“e disse:”, “and said:”): we have moved them into the verse they belong to in the Hebrew text.
In the public domain according to eBible.org and the Società Biblica in Italia. We are looking for an equally free Jewish Italian translation, such as the Pentateuco of Samuel David Luzzatto (1871–1876), of which no reliable digital text exists yet.
eBible.org: Riveduta 1927
English translation: King James Version
Status: Public domain
The 1611 English translation in the standard text of 1769, in the edition with Strong’s numbers distributed by eBible.org and the CrossWire Bible Society. Those numbers allow Toraoke to link each English phrase to the Hebrew word it translates (the link is computed by Toraoke). Words in italics were added by the translators.
In the public domain outside the United Kingdom, where printing remains subject to the Crown’s letters patent.
eBible.org: King James Version
Russian translation: Synodal
Status: Public domain
The Russian Synodal translation (Синодальный перевод), completed in 1876, from the RusSynodal module of the CrossWire Bible Society. The square brackets are those of the edition: they enclose the words added by the translators and the passages that the Synodal takes from the Septuagint and that are not in the Hebrew text.
In the public domain.
CrossWire: RusSynodal
Transliteration, names of the te’amim, explanations and summaries
Status: Made for Toraoke
The Latin transliteration under the words is computed by rules written for Toraoke from the vowels, dagesh and accents. The names of the te’amim and their explanations, the grammatical labels in Italian, the summaries of the sidrot and aliyot and the links between English phrases and Hebrew words are Toraoke’s. The links between the Italian and Russian words and the Hebrew ones (the underlining while listening or when hovering with the pointer) are also computed by Toraoke, but automatically, by a statistical model and a language model (LaBSE): they are right in about nine cases out of ten and can be wrong, especially in poetic passages and where the translation is free. The division into sidrot and aliyot follows the traditional custom of the weekly reading.
External links
Status: Link only
Each verse links to its page on Sefaria and on Blue Letter Bible; in the “Grammar” panel the lemma links to the Blue Letter Bible lexicon and to the Brown-Driver-Briggs dictionary on Sefaria; the cantillation panel links to the Wikipedia articles on the te’amim.
These are only links: Toraoke does not copy the content of these pages.
Sources used in the past
Status: Link only
Until November 2026 Toraoke took the Hebrew text, the grammatical analysis, the King James Version, the Synodal and the links between English and Hebrew from the pages of Blue Letter Bible, and the Italian translation from the Bibbia CEI 2008 (© Fondazione di Religione Santi Francesco d’Assisi e Caterina da Siena) on bibbiaedu.it. We have replaced them with the free sources listed above, because those sites do not authorize copying their content.
Voices and reading traditionsקְרִיאָה
Every recording carries two sets of rights, even when the text and the traditional melodies are free: those of the person who reads or chants and those of whoever made the recording. That is why a recording can be heard on Toraoke only after its rights have been checked; being downloadable elsewhere is not enough.
Recordings you can hear on Toraoke
Rav Dan Be’eri — Reading
Status: Declared license · authorization to be confirmed
A reading of the entire Tanakh in Hebrew, without cantillation, by Rav Dan Be’eri. On Toraoke: one clip per verse, with the words highlighted while listening.
- License
- CC BY-SA 3.0, declared on Internet Archive by whoever uploaded the recordings
- Attribution
- “Reading: Rav Dan Be’eri, Tanakh Narration. Source: Internet Archive. Declared license: CC BY-SA 3.0. Audio split into verses and re-encoded.”
- Changes
- each chapter is split into one clip per verse and re-encoded as MP3; the start of each word is estimated automatically
- Derived clips
- distributed under the same license, CC BY-SA 3.0
We do not yet have written confirmation from Rav Be’eri, or from whoever holds the rights to the recordings, that they were released under this license. If that is you, or you know who it is, write to us: we will correct the credits or withdraw the recordings on request.
Internet Archive · Genesis · Internet Archive · Exodus · Internet Archive · Leviticus · Internet Archive · Numbers · Internet Archive · Deuteronomy
Rav Dan Be’eri — Cantillation
Status: Declared license · authorization to be confirmed
The Tanakh chanted by Rav Dan Be’eri with the te’amim, in the Jerusalem Sephardi tradition (sefaradi-yerushalmi), in the simplified version used in schools, as Rav Mordecai Attiya taught it (as the source describes it). On Toraoke: one clip per verse, with the words highlighted while listening.
- License
- CC BY-SA 3.0, declared on Internet Archive by whoever uploaded the recordings
- Attribution
- “Cantillation: Rav Dan Be’eri, Tanakh Cantillation (Jerusalem Sephardi tradition). Source: Internet Archive. Declared license: CC BY-SA 3.0. Audio split into verses and re-encoded; word timings estimated automatically.”
- Changes
- each chapter is split into one clip per verse and re-encoded as MP3; the start of each word is estimated automatically
- Derived clips
- distributed under the same license, CC BY-SA 3.0
We do not yet have written confirmation from Rav Be’eri, or from whoever holds the rights to the recordings, that they were released under this license. If that is you, or you know who it is, write to us: we will correct the credits or withdraw the recordings on request.
Internet Archive · Genesis · Internet Archive · Exodus · Internet Archive · Leviticus · Internet Archive · Numbers · Internet Archive · Deuteronomy
Unpublished recordings
We thank and mention them because they are part of Toraoke’s history or of the tradition we want to make known, but they cannot be heard on Toraoke: mentioning them here does not mean having permission to use them.
Moise Levy — Italian-rite cantillation
Status: Not published
The Torah chanted according to the Italian rite in the voice of Moise Levy, published by Moise Levy Editore (Milan) together with its edition of the Torah with Rashi, and available to listen to on its website, libri.levy.it.
Earlier versions of Toraoke used excerpts from these recordings. In this version they are not published: the publisher’s terms require written authorization to republish the recordings, and we do not have it. To listen to them, see the publisher’s website.
Abraham Shmuelof — Reading
Status: Not published
A reading of the Hebrew Bible (Letteris text) by Father Abraham Shmuelof (died 1994), recording ℗ 1992 Talking Bibles International, available to listen to at listen.talkingbibles.org.
Earlier versions of Toraoke used excerpts from this recording, taken from videos published on YouTube, and wrongly attributed the voice to “M.L. Kohen”, who had made those videos: we apologize. The voice is Abraham Shmuelof’s, as confirmed by the author of the videos and by the descriptions that accompany them. In this version it is not published: we do not have the producer’s permission.
Historical Italian cantillation
There has been a Jewish presence in Italy for more than two thousand years, and its communities, of the Italian, Sephardi and Ashkenazi rites, have preserved their own ways of chanting the Torah, differing from city to city. Many of these melodies survive only in field recordings and in the memory of a few readers. Making them known to those who are learning, with the respect due to those who preserved them, is one of Toraoke’s aims.
Thesaurus of Italian Jewish Music — Centro Internazionale Leo Levi
Status: Not published · under study
The Thesaurus of Italian Jewish Music of the Centro Internazionale Leo Levi in Florence collects recordings of Torah readings in many Italian communities: the field recordings made by Leo Levi in the 1950s for the Centro Nazionale Studi di Musica Popolare, now in the Archivi di Etnomusicologia of the Accademia Nazionale di Santa Cecilia; those of Giovanni Marzi in Reggio Emilia (1954); and more recent recordings. Toraoke is studying 46 of them, from 14 communities: Ancona, Casale Monferrato, Ferrara, Firenze, Livorno, Mantova, Padova, Pitigliano, Reggio Emilia, Roma, Torino, Trieste, Venezia, Verona.
None of these recordings is published on Toraoke. Before making them available to listen to, we want to clarify for each one the date, the history of its publication and the consent of the readers or their heirs, in agreement with the Centro Leo Levi. The names of the individual readers will appear here when their recordings are published with their consent.
Accademia Nazionale di Santa Cecilia, Bibliomediateca
Technology
Speech recognition: zipformer2-66m-streaming-he
Status: Open license
The karaoke recognizes the reading with the zipformer2-66m-streaming-he model by Yakov Kolani, published by ivrit.ai and trained on Hebrew recordings from ivrit.ai; it derives from a model of the icefall project (k2-fsa, Apache 2.0 license). The model (85 MB) is downloaded into the browser and recognition takes place on the device: the voice does not leave the device. The model is used as published.
License CC BY 4.0.
The model on Hugging Face · ivrit.ai · icefall
ONNX Runtime Web 1.23.2
Status: Open license
The engine that runs the model in the browser (WebAssembly). © Microsoft Corporation.
MIT License: the full text is in the Licenses section.
onnxruntime.ai · Source code
Whisper and faster-whisper
Status: Used only in preparation
The Whisper speech recognition model (OpenAI) in its “small” version, run with faster-whisper (SYSTRAN), was used during preparation to estimate the start of each word in the recordings and to check the verse cuts. It is not part of the app; the word timings it estimated are.
MIT License for both.
Whisper · faster-whisper
Typefaces
Status: Open license
Alegreya and Alegreya Sans for Latin and Cyrillic text, Noto Serif Hebrew for Hebrew (בְּרֵאשִׁית בָּרָא). They are served by this site, without external services.
Alegreya: © Huerta Tipográfica (The Alegreya Project Authors), 2011 · source · license text
Alegreya Sans: © Huerta Tipográfica (The Alegreya Sans Project Authors), 2013 · source · license text
Noto Serif Hebrew: © The Noto Project Authors, 2022 · source · license text
SIL Open Font License 1.1. The typeface names remain their authors’.
Toraoke’s code
Status: Made for Toraoke
The pages, the karaoke (audio filters, running the model, following the reading), the transliteration and the preparation tools were written for Toraoke.
Licenses
The licenses of the materials Toraoke publishes, with the full text where the license requires it to be reproduced. This page does not replace the original licenses, which prevail, and is not legal advice.
| CC BY-SA 3.0 · legal code | Rav Dan Be’eri’s recordings and the derived clips |
| CC BY 4.0 · legal code | Lemmas and grammatical analysis (Open Scriptures Hebrew Bible); speech recognition model; Creative Commons texts in the video |
| SIL OFL 1.1 · legal code | Alegreya, Alegreya Sans, Noto Serif Hebrew |
| MIT | ONNX Runtime Web; Whisper and faster-whisper (preparation only) |
| Apache 2.0 | icefall, from which the recognition model derives |
ONNX Runtime Web license notice (MIT)
MIT License
Copyright (c) Microsoft Corporation
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
Corrections and permissions
If you hold rights to a material mentioned here, as author, performer, producer, publisher or heir, or if you find an error in a name, a date or a source, let us know: tell us the material and the page on which it appears. If a rights holder asks us to remove a material, we remove it while we verify; if they grant us its use on different terms, we update the credits as they ask.
The address for reports is in preparation and will be given here.